April 2008
Hilton Foundation
Project to End Homelessness for
People with Mental Illness
in Los Angeles:
Changes in Homelessness,
Supportive Housing, and Tenant Characteristics
Since 2005
Prepared for the Corporation for Supportive Housing
by Martha R. Burt, Ph.D., Urban Institute
Acknowledgements
This report was prepared for the Corporation for Supportive Housing by Martha R. Burt at the
Urban Institute. Its contents are the views of the author and do not necessarily reflect the views or
policies of the Corporation for Supportive Housing or the Urban Institute, its trustees, or funders.
Inquiries
If you are interested in learning more, please see www.csh.org for additional on-line resources and
materials, including information regarding the communities in which we currently work. If you have
questions or comments regarding this publication, please contact the CSH Resource Center at
info@csh.org. This publication is available to download for free at www.csh.org/publications.and at
www.urban.org.
The Corporation for Supportive Housing (CSH) is a national non-profit organization and Community Development Financial Institution
that helps communities create permanent housing with services to prevent and end homelessness. Founded in 1991, CSH advances its
www.csh.org
mission by providing advocacy, expertise, leadership, and financial resources to make it easier to create and operate supportive housing.
CSH seeks to help create an expanded supply of supportive housing for people, including single adults, families with children, and young
adults, who have extremely low-incomes, who have disabling conditions, and/or face other significant challenges that place them at on-going
risk of homelessness. For information regarding CSH’s current office locations, please see www.csh.org/contactus.
We encourage nonprofit organizations and government agencies to freely reproduce and share the information from CSH publications. The
organizations must cite CSH as the source and include a statement that the full document is posted on our website, www.csh.org. Permissions
requests from other types of organizations will be considered on a case-by-case basis; please forward these requests to info@csh.org.
Information provided in this publication is suggestive only and is not legal advice. Readers should consult their government program
representative and legal counsel for specific issues of concern and to receive proper legal opinion regarding any course of action.
© 2008 Corporation for Supportive Housing
TABLE OF CONTENTS
Introduction...............................................................................................................................2
Key Research Questions for This Report.........................................................................................2
Assumptions and Predicted Outcomes.............................................................................................3
Data Sources Used for This Report...........................................................................................4
Changes in Outcomes Between Baseline and 2007 ..................................................................5
Homeless, Chronically Homeless, and Mentally Ill Homeless Counts ........................................5
PSH Units in Los Angeles County in 2007 and 2004.....................................................................8
PSH Occupants in Los Angeles County...........................................................................................9
Homeless Single Adults and Homeless Families ....................................................................10
Homeless, Chronically Homeless, or Not Homeless.............................................................10
Disabilities Present Among Current PSH Tenants .......................................................................11
Pathways to PSH ................................................................................................................................13
Funding for PSH in Los Angeles.....................................................................................................14
Capital Funding............................................................................................................................14
Funding Sources for Operating PSH .......................................................................................15
Funding Sources for Providing PSH Supportive Services ....................................................16
PSH Partnerships, Collaborations, and Policies ............................................................................18
Collaborations and Partnerships ...............................................................................................19
Mission to Serve ..........................................................................................................................19
Project Policies.............................................................................................................................20
Conclusions ..................................................................................................................................20
Evidence for the Effectiveness of Housing and Employment Services for Chronically
Homeless People with Multiple Disabilities........................................................................... 21
The Skid Row Collaborative.............................................................................................................21
Documenting the SRC’s Effectiveness ....................................................................................22
Meeting Housing Retention Milestones...................................................................................22
LA's HOPE.........................................................................................................................................25
Housing and Employment Outcomes of LA's HOPE..........................................................25
The Future of the Skid Row Collaborative and LA's HOPE......................................................26
Conclusions .............................................................................................................................27
References................................................................................................................................29
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
1
INTRODUCTION
In 2005, the Hilton Foundation gave CSH a five-year grant to launch an initiative in Los Angeles
County to reduce the number of long-term homeless people, with a special focus on reducing
homelessness among people with serious mental illness. To promote these outcomes, CSH has
been using grant money to: fund predevelopment work on various permanent supportive housing
(PSH) projects, invest in building the capacity of supportive housing providers, work with public
officials and other key stakeholders in the county and selected cities (Los Angeles, Long Beach,
Santa Monica, and others) to stimulate increased commitment to PSH, support the Special Needs
Housing Alliance, and pursue other activities with City and County of Los Angeles officials and
agencies.
CSH contracted with the Urban Institute to help evaluate this initiative. A first evaluation report
(Burt 2005a) served two purposes - it presented an evaluation plan to address the research questions
posed by the Foundation, and documented baseline levels of the target population and activities in
Los Angeles County focused on ending long-term homelessness. The second evaluation report
examined the extent of systems change in the first two years of grant funding and the extent to
which it could be attributed to CSH activities under the grant or connected to it (Burt 2007a). This
third report is an update of the first one, focusing on changes in baseline levels of the target
population and PSH availability in Los Angeles County. A fourth report, expected in late summer
2008, will update the policy impact/systems change information with activities and their effects in
2007.
Key Research Questions for This Report
This evaluation is designed to answer a number of research questions, using either existing data or
data collected specifically for this evaluation. This report provides information to answer the first
three questions (the fourth report will focus on answers to the fourth question):
1. Have the efforts of CSH and their local partners led to an increase in the units of permanent
supportive housing in Los Angeles?
o How many PSH units exist in Los Angeles at baseline (end of 2004)?
o How many PSH units are being created in Los Angeles over the life of Hilton’s grant
(five year period)?
o How many PSH units are “in the pipeline”?
2. Who is being served by permanent supportive housing? Are chronically/street homeless
people getting into supportive housing units?
o What percentage of PSH units is occupied by people most in need of PSH (i.e. those
who are chronically homeless with disabling conditions, and especially those with serious
mental illness)?
o How did formerly homeless tenants get into PSH (i.e. from the streets, from shelters,
transitional housing programs, institutions, etc.)?
o Are there procedural and outreach practices that either inhibit or facilitate housing
receipt and stability for those who have been chronically homeless?
3. Has there been a reduction in the number of street homeless people in Los Angeles, or at
least of people with severe mental illness living on the streets?
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
2
4. How have state and/or local public agencies and funding streams made changes to better
accommodate the development and operation of permanent supportive housing units and
the services that tenants need to achieve stability?
o Are more, or different, public agencies and actors on board (e.g., mayors, agencies in
specific cities, new county council support, etc.)
o Are public agencies better coordinating their efforts to serve chronically/street homeless
people?
o Has any additional funding been committed at the local or state level to develop and
operate supportive housing (e.g., more funding in the same streams, new streams)?
o How have local agencies and providers been able to leverage additional state and federal
resources to deliver services to PSH tenants?
Assumptions and Predicted Outcomes
Based on the research questions just described, we summarize the outcomes of interest as:
1. Significantly increased numbers of PSH units in 2010, compared to 2005;
2. A higher proportion of PSH units occupied by the hardest-to-serve formerly homeless
people, especially those with serious mental illness;
3. Fewer people with serious mental illness in the county’s shelters and on its streets;
4. Increased public funding of PSH and other mechanisms to help long-term homeless people
with serious mental illness and others leave the streets, including more federal dollars,
more/new state and local dollars, greater leveraging of existing state and local dollars, and
more efficient and effective uses for existing resources; and
5. Increased coordination at every level of the system, among funding agencies and funding
mechanisms, providers, referral sources and pathways to PSH, and policy makers.
CSH has been working under its Hilton Foundation grant to increase (1) the number of public
agencies and public officials knowledgeable about long-term homelessness and committed to
working to end it; (2) the level of coordination among these agencies, and also among the
organizations able to develop PSH and offer supportive services to keep people in housing; and (3)
funding for PSH, and also for relevant homelessness prevention activities. Increases in these areas
should lead to more PSH, and also to better linkages among those with housing to offer and those
in contact with the neediest potential tenants. Changes in all the factors just described should result
in less long-term homelessness, including fewer people with mental illness who become and remain
homeless.
We graphically depict our assumptions about how CSH activities might influence the various
outcomes of interest in Figure 1. The arrows in Figure 1 show the relationships we expect to see as
CSH proceeds with its work under its Hilton grant. We have designed data collection and analysis
for this evaluation to assure that we have the information needed to evaluate whether these
predicted relationships have indeed come to pass. We will be able to see whether, at the end of the
grant, we observe lower levels of long-term homelessness and homelessness among people with
mental illness. We will also be able to identify the ways in which CSH activities stimulated public
agencies and actors, and PSH providers, to change in ways that contributed to reduced levels of
homelessness.
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
3
Figure 1: Predicted Relationships Among CSH Activities, Intermediate Steps and Changes, and
Reduction of Long-Term Homelessness
+
Public agencies
and actors working
to end long-term
homelessness
CSH Activities under Hilton Grant
+
+
_
Coordination
among public
agencies, housing
providers, and
service providers
+
_
_
Funding for PSH
Capital
Operations
Services
Funding for prevention
at institutional discharge
Number of
long-term
homeless
people
Number with
serious mental
illness
+
+
+
More PSH
More links to
get right people
into PSH
DATA SOURCES USED FOR THIS REPORT
Much of the information in this report describing Los Angeles agencies and their partners offering
PSH, PSH projects, units, and policies, and PSH tenants comes from reports on a large CSH
initiative called Taking Health Care Home (THCH). THCH began in 2003 with a grant from the
Robert Wood Johnson Foundation. Through THCH, CSH has worked with states and localities to
demonstrate how they can create supportive housing that ends homelessness for people with
chronic health conditions including mental illness, alcohol and chemical dependency, and
HIV/AIDS, and how that experience can be replicated on a national scale. THCH’s primary focus
was on systems change at the state and local levels. The goal was to create systems and government
infrastructure within state, county, and/or city governments to produce supportive housing in a
more integrated and coordinated way.
An evaluation of THCH has been in place from its inception. The evaluation has produced two
policy reports, one describing system changes in THCH communities during the initiative’s first two
years (Burt and Anderson 2006) and the second describing developments, expansions, and
extensions of early systems change work during the initiative’s last two years (Burt 2008c). Another
two reports focused on the details of PSH creation and availability in THCH communities and the
pipeline of projects stimulated by THCH (Burt 2005b and Burt 2008a). 1 As the first report on the
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
1
All reports are available on the CSH website (www.csh.org/resources/publications).
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
4
relied on Burt 2005b for much of its description of PSH in Los Angeles at baseline, so this report
relies on information from Burt (2008a) for its descriptions of PSH agencies, projects, and tenants in
Los Angeles three years later.
Other data used in this report include:
•
•
•
Street and shelter count data - the Los Angeles Homeless Services Authority (LAHSA)
conducted the county’s first countywide point-in-time homeless count in 2005 and repeated
it in 2007 (LAHSA 2005, 2007). LAHSA’s reports of results include information from
similar counts conducted in Long Beach, Glendale, and Pasadena.
Data on housing outcomes for chronically homeless people participating in the Skid Row
Collaborative, a federal demonstration project funded by HUD, the Department of Health
and Human Services, and the Department of Veterans Affairs, and being run by a
partnership of the Skid Row Housing Trust, Lamp Community, and JWCH Institute with
participation from the Los Angeles County Department of Mental Health (DMH), the
Housing Authority of the City of Los Angeles (HACLA), and the Greater Los Angeles
Veterans Healthcare System, reported in Burt (2007).
Data on housing and employment outcomes for chronically homeless people participating in
LA's HOPE, a federal demonstration project funded by HUD and DOL, and being run by
the City of Los Angeles’s Community Development Department (CDD) and its partners,
DMH, and HACLA, reported in Burt (2008b).
CHANGES IN OUTCOMES BETWEEN BASELINE AND 2007
Homeless, Chronically Homeless, and Mentally Ill Homeless Counts
In January 2005, LAHSA conducted the first-ever count of homeless people in Los Angeles County;
it repeated that count, with considerably improved methodology, in January 2007. Thanks to the
2005 survey, in the first report for this evaluation we reported a baseline number of homeless people
in the county in 2005, including the numbers of single adults, long-term homeless people, and
homeless people with serious mental illness, alone or with a co-occurring substance abuse disorder.
We can now also report the results of LAHSA’s 2007 survey. Methodological improvements for the
2007 effort were so substantial that we believe it may not be appropriate to interpret the differences
between some aspects of the 2005 and 2007 estimates or population characteristics as improvements
in the status of homelessness in Los Angeles County. Some of them at least, are probably the result
of methodological rather than substantive changes. So we report the two sets of results and discuss
what they might mean. LAHSA will conduct a third countywide survey in 2009, using the more
refined techniques applied in 2007. At that time we will have more confidence that major changes
in the level of homelessness and population characteristics between 2007 and 2009 reflect true
change, including differences in the number of long-term homeless people in the county and those
with serious mental illness.
Table 1 shows a 16 to 17 percent reduction in the estimate of people homeless during a single day in
late January between 2005 and 2007. For the LAHSA areas, some part of this decrease is probably
due to improvements in the design of the count, especially in the way that the researchers treated
census tracts where there was a very high probability of encountering homeless people (these were
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
5
known as “must enumerate” tracts). But a parallel and almost identical decline in Long Beach,
Pasadena, and Glendale, which have done their own counts and surveys for several years, indicates
that some part of the decline, and perhaps a large part, is real.
Table 1: Los Angeles County One-Day Homeless Counts,
2005 and 2007
Los Angeles County Totala
In LAHSA Continuum of Care (CoC) b
In Long Beach, Pasadena, or Glendale CoCs
Disabilities Experienced by Homeless People—LAHSA Survey Only
Serious Mental Illness
Chronic Substance Abuse Disorders
Physical Disabilities
At Least One Disability
Three or More Disabilities
Chronically Homeless—LAHSA Count Onlyc
As Proportion of All Homeless People
As Proportion of Single Adults
2005
88,300
82,300
6,000
2007
73,700
68,700
5,000
31%
43%
35%
46%
55%
35,000
43%
64%
31%
42%
35%
74%
57%
22,400
33%
49%
Percent Change
- 16.5
- 16.4
- 16.7
-36.0%
Sources: “2007 Greater Los Angeles Homeless Count,” accessed at www.lahsa.org on October 15, 2007, and “2005 Greater Los Angeles
Homeless Count,” accessed at www.lahsa.org on January 8, 2007.
a Rounded to nearest 100.
b Differences in estimates of population size between 2005 and 2007 must be interpreted with great caution, because changes in sampling
and weighting strategies greatly affected the 2007 results.
Data on population characteristics for the areas of the LAHSA survey come from interviews
conducted with homeless people shortly after the night of the count. Table 1 also shows that there
has been essentially no change in the prevalence of serious mental illness, chronic substance abuse
disorders, physical disabilities, or multiple disabilities among homeless people in Los Angels County
between 2005 and 2007. However, there does appear to be a substantial increase in the proportion
of homeless people reporting that they have at least one disability.
Table 1 also reveals a substantial drop in chronic homelessness as a proportion of the entire
homeless population (from 43 to 33 percent) and of single adults, who are the group to which the
“chronically homeless” definition applies (from 64 to 49 percent). Some of the same
methodological improvements that account in part for a lower overall count undoubtedly also
account for most of this reported decrease. Chronically homeless people are least likely to be found
in homeless shelters and most likely to be found in the “must enumerate” census tracts - these tracts
are identified in large part on the basis of their continuing presence. The two methodological
differences most likely to be contributing to reported changes in population characteristics between
2005 and 2007 are changes in how “must enumerate” census tracts were handled and changes in
where post-count interviews were conducted.
Because Los Angeles County is so huge and it would be impossible to cover every inch of it during a
three-night count, the LAHSA surveys take a sample of the 1,886 census tracts in its survey area and
go only to these tracts to do their counting. Selected census tracts were of two types - “must
enumerate” and “sampled.” A “must enumerate” tract represented only itself; 235 “must
enumerate” tracts were identified and included in the 2007 count. Each “sampled” tract, of which
there were 270 in 2007, represented about 6 of the remaining 1,651 tracts, on average.
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
6
When researchers developed their 2005 and 2007 estimates of the total number of homeless people
in Los Angeles County, they used a procedure known as “weighting,” which “blows up” the actual
count in the 270 sampled tracts to make them represent all 1,651 tracts from which they were
selected. The results of the actual count - the number of homeless people actually seen in a sampled
tract - will then be multiplied by the weight, which for illustrative purposes we will say is 6, to
estimate the total number of homeless people in the several tracts represented by the sampled tract.
So if searchers saw 10 people on the night of the count, these 10 people would be “blown up” to 60
people for the final estimate. A “must enumerate” tract receives a weight of 1, so when one
multiplies the number of people observed by the weight (1), you get just the number of people
observed. Thus if searchers saw 375 homeless people in a “must enumerate” tract, that tract would
contribute 375 people to the final estimate.
It is easy to see that it would make a big difference to the final count if researchers failed to identify
a census tract with large numbers of homeless people as a “must enumerate” tract but left it in the
group of tracts from which random tracts were sampled. Then that hypothetical tract with 375
observed homeless people and a presumed weight of 6 would contribute not 375 but 2,250
homeless people to the count, and a very large proportion of them would be chronically homeless
people. This is what happened in 2005, when the procedures for identifying “must enumerate” tracts
were far less well developed than they were for the 2007 count. In 2005, several tracts that each
included hundreds of homeless people were found among the sampled tracts, having not been
identified as “must enumerate” tracts. Therefore their hundreds of people were “blown up” in the
manner described above, becoming thousands of homeless people during the process of developing
a final estimate rather than just being counted as themselves.
Data on people’s length of homelessness and presence of disabilities - the two factors that together
determine whether the person should be considered chronically homeless - come from the postcount interviews. In 2005 these interviews were conducted only in shelters and in “must
enumerate” census tracts, but not in randomly selected tracts. Further, they were not allocated in
relation to the proportion of homeless people found in each type of location. The information they
yielded on population characteristics cannot therefore be considered representative of the whole
population of people homeless in Los Angeles in January 2005. In 2007, interviews were conducted
in every type of tract and shelter, and were allocated in proportion to where people were counted.
The information collected in 2007 is thus more likely to provide an accurate picture of population
characteristics than the 2005 approach.
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
7
Figure 2: Location of Chronically Homeless Single Adults,
by Service Planning Area, 2005 (green) and 2007 (blue)
9,000
8,336
8,295
8,000
7,000
6,269
6,000
5,056
5,000
3,626
4,000
3,664
3,593
2,906
3,000
2,106
2,025
2,515
1,935
2,000
1,000
2,555
2,451
1,338
593
0
SPA 1Antelope
Valley
SPA 2-San SPA 3-San
Fernando
Gabriel
Valley
Valley
SPA 4SPA 5-West
SPA 6SPA 7-East
SPA 8Metro Los
Los
South Los
Los
South
Angeles
Angeles
Angeles
Angeles Bay/Harbor
The effects of these methodological changes can be seen dramatically in figure 2, which shows the
distribution of chronically homeless single adults by Service Planning Area (SPA). In addition to
showing that chronic homelessness is not evenly distributed among the county’s eight SPAs, figure 2
also shows substantial drops in estimates of chronic homelessness between 2005 (green columns)
and 2007 (blue columns) for every SPA except SPA 4 - the SPA containing Skid Row in downtown
Los Angeles. Of all the possible census tracts in Los Angeles County that might have been
identified as “must enumerate,” in 2005, researchers were least likely to miss those in downtown Los
Angeles. Since almost all the same tracts in SPA 4 were identified as “must enumerate” in both
years, the estimates of chronic homelessness did not change. In the remaining seven SPAs, better
identification of “must enumerate” tracts and more appropriate allocation of post-count interviews
contributed to the reduced estimates of chronic homelessness.
PSH Units in Los Angeles County in 2007 and 2004
For CSH’s Taking Health Care Home (THCH) initiative, an evaluation estimated the number of
PSH units in Los Angeles County in 2004 and 2007 (Burt 2008a). We use the results of surveys in
these two years as the baseline for the Hilton project and as an indication of progress in the project’s
first three years. (We will repeat this survey in Los Angeles in 2009 to provide data on PSH status at
the end of the Hilton project.)
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
8
5,455 PSH units
2004 estimate, based on surveys of 38 PSH projects in Los Angeles County,
weighted to represent all PSH in the county. 2
6,030 PSH units
2007 estimate, based on surveys of 45 PSH projects in Los Angeles County,
weighted to represent all PSH in the county.
2,200 PSH units
Pipeline for the future - at least this many PSH units are currently in
development.
Comparing the count of chronically homeless single adults - 35,000 in 2004 and 22,000 in 2007 - to
the number of PSH units available in those years gives some idea of the size of the gap that must be
filled before the county will reach the goal of eliminating chronic homelessness. Almost 600 new
units opened and housed formerly homeless tenants in those years; pipeline figures show that at
least another 2,200 are actively in development and should come on line within the next one to three
years. These 2,200 pipeline units are ones in which the CSH Los Angeles office has some
involvement, using Hilton Foundation and other resources to stimulate development. Others are
probably also on the drawing boards, but without CSH involvement. Even so, to reach the goal, the
county will need about 14,000 more PSH units or other strategies for ending chronic homelessness.
Los Angeles’s existing PSH projects are mostly single-site - 88 percent in 2007 and 91 percent in
2004. Most of these projects (82 percent in 2007 and 78 percent in 2004) are in buildings
completely occupied by formerly homeless individuals - what we call “dedicated” buildings. The rest
are in buildings whose tenants include some never homeless people, or at least people who did not
come to the units directly from homelessness. We refer to this latter configuration as “mixed use,”
indicating that it accommodates both formerly homeless and never homeless individuals and
families. Non-PSH units in these mixed use buildings are almost always “affordable,” meaning that
tenants pay no more than 30 percent of their income toward rent. Adding units in mixed-use
buildings occupied by never homeless people to the PSH units already noted brings the estimated
total number of units in buildings used for PSH up to about 6,300 in 2007. Only about 9 to 12
percent of PSH projects used a scattered site configuration - these were mostly units associated with
agencies using McKinney-Vento Shelter Plus Care certificates to help their clients obtain apartments
in the community.
PSH Occupants in Los Angeles County
THCH project surveys asked about the types of people occupying PSH units, and where those
people were staying immediately before moving into PSH.
2
In 2004 we received completed surveys from 38 projects offering about 2,100 beds; in 2007 we received completed
surveys from 45 projects offering about 2,250 PSH beds. Responding projects were sampled, according to their size,
from the much larger number of projects in Los Angeles that were open and occupied at the time of the surveys and
offered PSH for formerly homeless people with disabilities (118 in 2004 and 142 in 2007). Each sampled project
represented a stratum defined by project size—1-9 units, 10-19 units, etc. To get these estimates of PSH units, as well
as those for all the tables in this report based on THCH, we weighted the data from the surveys we received so they
represent all PSH beds, units, and projects in Los Angeles County. Weights for projects were calculated using the
following formula: (number of beds in a stratum/number of beds in the project)/number of projects in the sample
from that stratum.
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
9
Homeless Single Adults and Homeless Families
For the first time in 2007, the survey asked about the household status of PSH occupants. Los
Angeles projects reported that 89 percent of their units were occupied by single adults and 5 percent
were occupied by families (an adult with at least one child present). Household status was not
known or not reported for 6 percent of the units.
Homeless, Chronically Homeless, or Not Homeless
The most important question, from the point of view of ending chronic homelessness, is whether
current PSH occupants had been chronically homeless before moving into PSH, as one would hope
that PSH is being reserved for people whose long-term homelessness indicates limited ability to
obtain or retain housing without the substantial assistance offered by PSH. The results for open and
occupied PSH units in Los Angeles County for which former homeless status is known are:
•
32 percent in 2007 and 22 percent in 2004 were occupied by people who were chronically
homeless at move-in.
•
62 percent in 2007 and 54 percent in 2004 were occupied by formerly homeless people
whose pattern of homelessness at move-in did not qualify them as chronically homeless.
•
6 percent in 2007 versus 10 percent in 2004 were occupied by people who were never
homeless, including low-income people in need of affordable housing, ex-offenders, and
people at high risk of homelessness.
•
Projects were more knowledgeable about their tenants’ homeless status in 2007 than they
had been in 2004, but the increased information did not significantly change the proportion
reported in each category (chronically homeless, other homeless, not homeless).
Looking at all PSH in Los Angeles County together, we can see a slight increase in the proportion of
all PSH units occupied by people who had been chronically homeless. This statistic may not reflect
recent developments, however, because such a large proportion (90 percent) of all PSH beds in 2007
were already in existence in 2004 and the characteristics of occupants presumably have not changed
very much. Older projects with established eligibility requirements are probably replacing departing
tenants with others with similar characteristics. Therefore we compared the proportion of beds
opened after 2004 that were occupied by formerly chronically homeless people to the same proportion
in beds that were already open in 2004. If the recent policy push toward targeting new beds on the
longest-term homeless people is having an effect, we should see it in beds that came on line after the
baseline 2004 survey.
These figures tell a very different story - 65 percent of new PSH beds were occupied by chronically
homeless people, compared to only 26 percent of old PSH beds. Figure 3 displays these differences,
which indicate that agencies developing new PSH projects are heeding the policy priorities of
housing chronically homeless people.
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
10
Figure 3: “New” Beds Are More Targeted to Chronically Homeless People
70
65
Percent of Beds
60
50
40
30
26
20
10
0
"Old" Beds
"New" Beds
Disabilities Present Among Current PSH Tenants
Another critical question about PSH occupants is what disabilities they have. It may be that many,
even though they do not have a history of chronic homelessness, would be highly likely to become
and remain homeless without the supports offered by PSH. People with serious mental illness
might fall into this category, as might some people with chronic substance abuse, HIV/AIDS,
physical disabilities, or chronic health problems.
Table 2 shows what we learned about Los Angeles PSH tenants from THCH project surveys in
2004 and 2007. The first set of columns shows the average proportion of all units in PSH projects
in Los Angeles County that are occupied by people with particular disabilities. 3 These figures are
not mutually exclusive, so tenants may be counted in more than one category. It is clear that in both
years, serious mental illness was the disability most common among PSH tenants, and that a
significant proportion of tenants also had problems with addictions. We know that PSH is intended
to serve people with disabilities, so we would expect that every tenant has at least one disability. It is
possible that the same person was counted under alcohol abuse and drug abuse, and someone could
have been counted as having HIV/AIDS or physical disabilities and any of the other categories. As
we saw in the LAHSA data reported above, the majority of homeless individuals have three or more
such disabling conditions.
3
This analysis concentrates on projects, showing the average proportion of beds per project that are occupied by people with various
disabilities. This approach treats similarly a project with 10 beds and a project with 100 beds that both say 100 percent of their beds
are occupied by people with serious mental illness – both reported 100 percent regardless of the number of beds involved. Another
way of thinking about this analysis would be to focus on beds, asking what proportion of all the PSH beds in a community are
occupied by people with various disabilities. We chose to focus on projects because we were concerned with project policies and
whether project specialties or exclusionary policies resulted in increased or decreased access to PSH for people with particular
disabilities. We also ran the analysis using beds rather than projects – for the most part the two analyses tell the same story. The
differences worth noting include (1) the percent of beds dedicated to tenants with serious mental illness, which was slightly smaller
than the analysis illustrated in table 2 (by around 6-7 percentage points in both 2004 and 2007), and (2) the change in the percent of
beds occupied by tenants with a dual diagnosis between 2004 and 2007, which was also smaller than the analysis illustrated in table
2 (a 4 percentage point change rather than the 8 percentage point change).
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
11
Table 2: Los Angeles PSH Project Information
Regarding Tenant Disabilities
Disability
Serious Mental Illness
Alcohol Abuse
Drug Abuse
Dual Diagnosis
HIV/AIDS
Physical Disabilities
Average Proportion of
Units Occupied by People
with this Disability,
Across All Projects
2004
2007
55
67
28
29
49
51
23
41
15
12
8
9
Percent of Projects with
No Units Occupied by
People with this
Disability
2004
2007
2
1
15
27
29
11
4
8
35
50
25
4
Percent of Projects
whose Funding Requires
that they Serve People
with this Disability
2004
2007
31
32
18
17
18
17
20
17
20
16
14
3
The second set of columns in table 2 shows the proportion of projects that do not have any tenants
with each type of disability. These proportions in several categories increased considerably between
2004 and 2007. One might read these differences as indicating that PSH projects have gotten more
specialized in the years between the two surveys, but we think the more likely explanation is that for
2004 only 9 agencies participated in the survey, and most of them were large agencies that developed
large hotel conversion projects. Thus we had surveys for 38 projects but most were very similar to
each other and, being large, tended to have a relatively diverse tenant base. For 2007 we made a
point of including a greater number of smaller agencies and agencies that ran smaller projects, so we
received surveys from 13 agencies and 45 projects. The smaller projects are more specialized than
the larger ones, with the result that a higher proportion of projects in 2007 reported having no
tenants with certain disabilities. Missing data from certain projects is another factor that may be
influencing these results. Several of the “new since 2004” small projects did not report client
characteristics. We also did not get surveys back in 2007 from one agency with an HIV/AIDS
specialization that contributed two project surveys in 2004, which accounts for much of the change
in units and projects whose tenants do not have HIV/AIDS and/or physical disabilities.
The final set of columns in table 2 indicates one of the reasons why PSH projects serve people with
certain disabilities - these are the populations they were designed to serve, and most have funding
that requires them to serve people with these disabilities. People with serious mental illness are the
most “supported,” with 31 to 32 percent of projects having funding that requires they be served.
Physical disabilities are the least-supported, in the sense that project funding does not require that
they specifically be served. The prevalence of mental illness and substance abuse among the
population requiring PSH is probably the factor driving the proportion of units occupied by people
with these disabilities, however, as occupancy rates mostly exceed what might be expected based
solely on funding requirements.
From the data in table 2 we can conclude that even if current PSH projects are not occupied
primarily by people who were once chronically homeless, they are mostly occupied by people who
have homeless histories, albeit not long-term homelessness, and whose disabilities and conditions
would place them at high risk for long-term homelessness if PSH were not providing them with
stable housing.
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
12
Pathways to PSH
There is much controversy about the right recruitment strategies for PSH. The “housing first”
approach argues that PSH should be recruiting tenants directly from the streets or emergency
shelters, without requiring a stay in transitional housing first. This argument rests on the belief that
it is impossible to address the many clinical issues of chronically homeless people unless they are
first in stable housing situations. Other approaches argue that chronically homeless people are not
ready for, nor are they capable of sustaining, permanent housing until they have addressed their
clinical issues, stabilizing their mental illness through medication and gaining some control over their
substance use. In the world of mental health services, transitional housing programs are one
possible venue for acquiring this stability and becoming what providers consider “housing ready.”
In the homeless services world, however, transitional housing programs more usually function as
places where people acquire employment or parenting skills that will let them function in housing on
their own, without need of the continuing supportive services one finds in PSH.
Many Los Angeles PSH projects use a combination of these approaches, as table 3 shows. Taking
PSH tenants from transitional housing projects is most common, with only 10 percent of projects in
2007 having no tenants who came to them directly from transitional housing (up from none in
2004). Direct tenancy from streets or emergency shelters is the route taken by 35 percent of PSH
occupants, with 84 to 89 percent of PSH projects having some tenants who came to them directly
from these venues. Other venues include jails, hospitals, and residential treatment facilities. Finally,
it appears that in 2007, projects did not, and perhaps could not, report the route to PSH taken by 26
percent of their tenants - up from only 7 percent in 2004. Many more projects in 2004 were able to
report their tenants’ pathways to PSH, as only 7 percent of projects had tenants whose prior living
situation was reported as unknown. In 2007, only 38 percent of projects reported tenant pathways
for all their tenants.
Table 3: Los Angeles PSH Project Information
Regarding Tenant Living Situation Immediately Before PSH
Disability
The Streets
Emergency Shelters
Either the Streets or
Emergency Shelters
Transitional Housing
Other Venues
Don’t Know
Percentage of Units Across All
Projects Occupied by People who
Came from this Location
2004
2007
20
12
15
23
35
24
33
7
35
28
11
26
Percent of Projects with No Units
Occupied by People who came from
this Location
2004
2007
13
15
19
21
11
0
9
92
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
16
10
66
38
13
Funding for PSH in Los Angeles
Data on PSH funding in Los Angeles come from THCH (Burt 2008a) and was weighted as
explained in footnote 2. Along with projects in other THCH communities, each Los Angeles
project sampled in 2007 completing a survey was asked to report the sources of capital, operating,
and services funding for the past year. In the case of capital resources, many projects were able to
go back to their original development funding. We report capital, operating, and services funding
separately, looking for the largest/most important sources, average number of sources per project,
and total dollar amounts. Total dollar amounts for capital go back to project beginnings, while total
dollar amounts for operating and services are just for the reporting year. For surveys completed in
2007 this would most often be calendar year 2006 or fiscal year 2007. In what follows, we present
information for projects in the 2007 THCH sample for Los Angeles and compare it to parallel data
collected in 2004. For more detail on funding sources and levels in other THCH communities,
please see Burt (2008a).
Capital Funding
Capital funding data are available from Los Angeles THCH surveys for 35 of 43 2007 projects and
31 of 37 2004 projects that had development costs. Los Angeles projects identified 22 sources of
operating funds in 2007 and 14 in 2004, including foundations, congregations, businesses, and
miscellaneous public sources (e.g., CDBG, and state or local rental assistance programs). In 2007,
these projects reported receiving a total of $558.9 million in capital investment since their inception.
In 2004, projects reported $395.8 million in capital funding. About 40 percent of this difference is
accounted for by capital for projects that opened since 2004; better data collection in 2007 is likely
to account for most of the rest. Projects reported an average of 4.5 different capital funding sources
in 2007 (4.7 different sources for 2004).
Per-unit costs for capital construction in Los Angeles come from the surveys received in 2007,
which we judge to most completely account for capital costs. For all PSH units involving capital
investment that were open in 2007, the average per-unit capital used in their construction or
rehabilitation was about $97,400. Since units open in 2004 were often constructed many years ago
under less expensive construction circumstances, we also calculated per-unit costs for units open in
2004 and those that opened after 2004. These costs are: $90,500 for units open in 2004, and
$160,000 for units that opened in 2005, 2006, or 2007.
Table 4 reveals the major sources of capital funding for PSH in Los Angeles in 2004 and 2007.
LIHTC is far and away the biggest contributor (41–43 percent of all capital funding). The
Redevelopment Authority contributed 17–19 percent and housing finance agencies contributed 6–9
percent of capital funding for PSH in Los Angeles. These three biggest capital funding sources
contribute 65 to 69 percent of all capital resources. Housing trust funds, CDBG, and commercial
banks also make important contributions to PSH in Los Angeles.
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
14
Table 4: Comparing Proportional Contributions of Different Capital Sources in
Los Angeles in 2004 and 2007
Sources
LIHTC
Redevelopment Agency
Housing Finance Agency
Housing Trust Fund
CDBG
Commercial Bank
Other Local
Mainstream Service Agenciesa
HOME
Public Housing Authorities (including Section 811)
McKinney-Vento (SHP and SRO Mod Rehab)
Other/Other Private
Federal Home Loan Bank
Total Dollars from Major Sources (Weighted, Millions)
Proportion from Major Sources
2004
41.2
19.3
8.6
8.0
4.6
4.7
4.4
0.7
1.9
3.4
0.9
1.1
1.0
2007
42.8
16.5
5.5
6.4
5.6
4.7
3.9
3.6
3.5
<0.1
1.6
1.4
0.6
$395.5
$555.5
99.9
96.2
aMainstream
agencies reported as contributing capital funds included mental health and housing and
community development health agencies.
Funding Sources for Operating PSH
In 2007, PSH projects in Los Angeles reported receiving $28.6 million in operating resources during
their most recent project year, which in most cases would have been calendar year 2006 or FY 2007.
This translates into approximately $7,000 per unit per year ($28.6 million/4,063 units, including
scattered-site units, in 35 projects). This is a considerable increase in reported operating revenues
compared to 2004, for which $25.7 million were reported (about $5,700 per unit per year). This 11
percent increase corresponds pretty closely to the 12 percent increase in PSH beds during the same
period.
Los Angeles projects identified 14 sources of operating funds in 2007 and 12 in 2004, including
foundations, congregations, businesses, and miscellaneous public sources (e.g., HOPWA, CDBG,
and state or local rental assistance programs). Projects reported an average of 2.5 sources for 2007
and 2.2 for 2004. Table 5 shows the proportional contribution of major sources of operating funds
for PSH projects in Los Angeles in 2007, 2004, or both.
In 2007, the single biggest source of operating revenues for PSH in Los Angeles was Shelter Plus
Care, a McKinney-Vento program. Shelter Plus Care contributed 46 percent, a substantial increase
from the 38 percent reported for 2004. Some small Los Angeles programs have their own Shelter
Plus Care grants, but the bulk of Shelter Plus Care in Los Angeles comes through the Department
of Mental Health (DMH). DMH routinely applies for additional Shelter Plus Care certificates every
year, and has increased its pool from 338 in 2004 to 464 in 2007, making this a substantial housing
subsidy resources for homeless people with serious mental illness.
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
15
Table 5: Comparing Proportional Contributions of Operating Revenue Sources
in Los Angeles in 2004 and 2007
Source
Shelter Plus Care
Tenant Payments
McKinney-Vento (SHP and SRO Mod Rehab)
Section 8
Commercial Rents
Special State or Local Appropriations
Redevelopment Agency
Mainstream Service Agenciesa
Total Dollars from Major Sources (Weighted, Millions)
Proportion from Major Sources
Los Angeles
2004
2007
38.3
46.3
30.2
30.9
12.3
11.5
4.2
3.8
3.4
2.4
-1.5
5.2
0.6
4.7
0.3
$25.1
98.3
$27.8
97.2
aThe mainstream agency reported as contributing operating funds was the mental health agency.
Tenant contributions to rent were next most important in both 2007 (31 percent) and 2004 (30
percent). The McKinney-Vento SHP and SRO Moderate Rehabilitation programs contributed
about 12 percent in both years. Section 8 subsidy certificates supported about 4 percent of PSH
units in 2004 and 2007. Some of the big residential hotel providers have many more units supported
by Section 8 than these figures reflect, but these units are not necessarily occupied by formerly
homeless people nor do they have significant amounts of services attached to them. They fall into
the category of “affordable” units rather than PSH.
The information available to describe operating revenues is a bit more confusing than for capital
sources, because projects sometimes gave a funding program (e.g., Section 8) and sometimes a
funding agency (e.g., public housing authority) in their survey responses. However, it is clear that
federal housing resources contribute a great deal to PSH operating revenues. In this category we
include Shelter Plus Care, McKinney-Vento (SHP and SRO Mod Rehab), and Section 8. In Los
Angeles, these sources taken together invested about $17.6 million in the 2006-2007 operations of
PSH projects - an amount equal to about 62 percent of all operating revenues reported.
Corresponding figures for 2004 for Los Angeles were $13.6 million and 55 percent.
Funding Sources for Providing PSH Supportive Services
The final funding component of PSH is the resources available for supportive services. Before
commenting on levels or changes in services funding, a few caveats about the data are important to
understand. For capital resources, some projects were not able to attach dollar figures to every
source they reported. For the most part these lapses were random rather than systematic, and are to
be expected. For services, however, (and a somewhat lesser extent for operations), the extent of
missing information is sufficiently important and potentially large enough to skew anything we
might say about access to services dollars.
We decided to compensate for this lack of information related to Medicaid funding where we could,
to get a better picture of services funding for 2004. We made some assumptions about Medicaid
funding for projects that were in both surveys and reported Medicaid funding in one year but not
the other. If funding was reported for 2004 but not for 2007, we attributed the 2004 amount to the
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
16
same project in 2007, augmented by the average increase of 1.03 for the same funding source that
we calculated from projects for which we did have two years of data. If the funding was reported
for 2007 but not for 2004, we attributed the 2007 amount to the same project in 2004, diminished
by the same average difference of 1.03. After doing this, we had service funding information for 58
percent of projects submitting a survey in 2007 and 82 percent of projects reporting in 2004. 4
In 2007, 26 PSH projects in Los Angeles, representing (after weighting) about 2,500 units, reported
receiving $8.6 million in supportive services funding resources during their most recent project year,
which in most cases would have been 2006 or FY 2007. This translates into approximately $3,500
per unit (including scattered site units) per year. The corresponding per-unit supportive services
estimated cost for 2004 was $4,000, as reported by 30 projects representing almost 3,900 units.
Table 6 shows the nine biggest sources of funding for supportive services in the reporting PSH
projects. First, total reported funding for services was down a substantial 45 percent between 2004
and 2007. The end of federal funding for the Skid Row Collaborative, one of 11 federal Chronic
Homeless Initiative demonstration projects, accounts for part of this decline. Ideally this funding
would have been replaced by local service resources; but this has not happened to date.
Table 6: Comparing Proportional Contributions of Supportive Services Funding in
Los Angeles in 2004 and 2007
McKinney-Vento (SHP and SRO Mod Rehab)
Medicaid
Mainstream Service Agenciesa
Tenant Payments
Other Private
Special State or Local Appropriations
HOPWA/Ryan White
Federal Chronic Homeless Initiative Funding
Veterans Affairs
Total Dollars from Major Sources (Weighted, Millions)
Proportion from major sources
2004
14.0
4.6
47.3
2.1
1.6
0.2
11.3
15.3
1.0
$15.6
97.4
2007
57.0
9.5
9.1
7.5
5.3
2.6
0.6
--$8.6
91.7
a Mental health agencies were the mainstream agencies contributing services funds to PSH projects in Los Angeles.
More important is greatly reduced contributions that flow through mental health agencies.
Reductions in revenues from state programs account for this difference. In 2004, 19 projects
reported receiving about $7.5 million in services funding from mental health agencies (not counting
funding for the Skid Row Collaborative through the federal Chronic Homeless Initiative). This
funding came largely through the California Department of Mental Health, and consisted of funding
from the Supportive Housing Initiative Act (SHIA) and one or two other state funding streams. All
such funding had ended in 2006, either because the state did not renew the program (SHIA) or
because it changed its approach to paying for services in ways that resulted in withdrawal of
supportive services from PSH projects. It is clear that projects have not been able to find alternative
funding sources to compensate for the loss of these state funds.
4
For operating dollars, 82 percent of Los Angeles projects submitting a survey in 2007 reported at least some
funding. For 2004, 74 percent of surveys received contained information on at least some operating resources.
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
17
The combination of fewer resources flowing through mental health agencies and the end of Chronic
Homeless Initiative funding would have raised the proportion of services funding attributable to
McKinney-Vento programs in any case, but it is also true that the amount of this type of funding
reported by surveyed projects more than doubled between 2004 and 2007, from $2.2 million to $5.4
million.
Finally, the Los Angeles County Department of Mental Health reported an increase between 2004
and 2007 of about $200,000 to provide services to tenants using DMH Shelter Plus Care
Certificates. During those years DMH continued to apply for Shelter Plus Care resources to house
more of its homeless clients - the number of certificates under DMH control grew from 338 in 2004
to 464 in 2007. A good part of the services funding increase is linked to providing services for these
additional tenants, but there is also a per-tenant increase of about 4 percent. Even with this small
increase, the service resources DMH reports for these Shelter Plus Care tenants are extraordinarily
low, at about $1540 a unit in 2007 compared to about $1480 in 2004. In 2007, Medi-Cal paid for
about 70 percent of these services with DMH making up the difference; in 2004 Medi-Cal covered
about 67 percent of the costs.
PSH Partnerships, Collaborations, and Policies
We wanted to understand the history, motivations, and policies of agencies offering PSH and the
projects they sponsor, and how they might have changed between 2004 and 2007, including:
•
The odds that the agencies currently involved in PSH will continue to help develop more
PSH units;
•
The collaborations and partnerships that agencies develop to support one or more aspects
of PSH—development, operations/management, and services;
•
What types of people they are likely to serve; and
•
Any policies that make it more or less likely that PSH will attract and retain tenants who
were once homeless.
Most agencies offering PSH in Los Angeles see themselves as “in it for the long haul.” They have
been involved in PSH in the past, are involved now, and expect to be involved in the future. In
2007, 44 percent of agencies were involved in all aspects of PSH, including development, operations,
and service delivery. This proportion is an increase from the 34 percent of agencies that were so
extensively involved in 2004; these agencies may be less inclined to enter into collaborations, as they
see themselves as covering the necessary components of PSH.
Another 47 percent of agencies in 2007 did two out of three PSH activities - most of these (37
percent) did the development and building management/operations aspects and partnered with
other agencies to provide services, but about 10 percent did the management/operations and service
delivery and had either partnered with another agency that did the development or ran scattered-site
projects that needed no development. Finally, by 2007 only a few agencies (5 percent) were new to
PSH or had no future intentions, whereas in 2004 about 1 in 4 agencies (27 percent) were not sure
of their commitment to continue producing and running PSH projects.
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
18
Collaborations and Partnerships
We gathered information about collaborations and partnerships from the agencies offering PSH and
also from the projects that actually house its tenants. Neither source (agencies or projects) reported
much change between 2004 and 2007 in their collaborations and partnerships. In 2007, fewer than
half (6 of 13) of the agencies surveyed named at least one partner for development, operations, or
services, whereas in 2004 7 of 9 agencies named such a partner. However, the average number of
partners per agency increased slightly at 1.8 development partners in 2007 (1.4 in 2004), 0.5
operations partners in 2007 (0.4 in 2004), and 1.3 service partners in each year (0.7 in 2004).
Projects were more likely than agencies to name partners in helping to support PSH tenants. In
counting an organization as a "partner," we erred on the side of caution. If it appeared that the
service provider had a significant commitment to serve PSH tenants, we counted that provider as a
partner. But if it was apparent that the PSH project simply referred tenants to various places as the
need arose, with no special commitment to serve on the part of the service provider, we did not
count those providers as partners. Many PSH projects named the same collaborators, such that in
2007 the 45 projects for which we have surveys named 41 partner organizations, which received a
total of 196 mentions. In 2004, the 38 projects with surveys named 26 partners and gave them a
total of 144 mentions.
Of the 45 projects with surveys in 2007, two did not name any partners. Three named a partner for
the development aspect of their projects, and all three named the same partner - A Community of
Friends (ACOF). Three 2007 projects were scattered-site, and thus had no development activity.
The remaining projects (37) were developed by the same agency that still ran them. In 2004, only
one PSH project named a development partner.
Partnering for the operations aspect of PSH was a bit more common in 2007, with 12 projects
naming operating partners. Eight of these, all developed by ACOF, used the John Stewart Company
for operations. Four, all developed by Homes for Life Foundation, worked with a different
operations management organization. Finally, the Los Angeles County Department of Mental
Health relied on mental health service providers throughout the county to handle the
operations/management aspects of housing subsidized through its Shelter Plus Care certificates. In
2004, 10 PSH projects named an operations partner, including one partner in addition to those
named in 2007.
Most projects (41 of 45) named at least one partner for services in 2007; on average, projects named
5.1 partners that worked with project tenants to deliver one or more types of service. Partners
receiving 15 or more mentions included the Los Angeles County Department of Mental Health and
its various local centers and clinics, Chrysalis (for employment), the VA, and the JWCH Institute
(for health care). In 2004, 32 of the 38 projects with surveys named at least one service partner, with
patterns of collaboration very similar to what still existed in 2007. The average number of partners
named was somewhat lower in 2004 than in 2007, at 3.8.
Mission to Serve
Between 2004 and 2007, the already high proportion of Los Angeles PSH projects that reported
having a mission to serve people with serious mental illness went from 93 to 97 percent, while those
with a mission to serve people with alcohol or drug addictions went from 59 to 52 percent. In Los
Angeles the primary commitment of most PSH projects appears to be to those with a serious mental
illness, as can be seen when one asks whether projects have a mission to serve those with mental
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
19
illness but not with addictions, those with addictions but not with mental illness, or those with both
serious mental illness and a co-occurring substance abuse disorder. Hardly any PSH projects have a
primary focus on substance abuse in the absence of a co-occurring mental illness - only 2 percent in
2007 and none in 2004. On the other hand, a substantial proportion of projects reported a mission
to serve only people with a serious mental illness who do not also have a co-occurring substance
abuse disorder - although more in 2007 than in 2004 (46 versus 34 percent).
Project Policies
We asked about a variety of project policies, from leasing arrangements to requirements that tenants
have a representative payee to what might cause a project to evict a tenant.
•
In the vast majority of units in Los Angeles PSH projects, tenants hold their own lease (92
percent in 2004 and 96 percent in 2007). Otherwise the lease is either jointly held with the
project or the project subleases to the tenant. In only 1 percent of units in either year does
the tenant live in a unit without a lease of any type.
•
No Los Angeles PSH project in either 2004 or 2007 requires tenants to have a representative
payee as a condition of rental. Nevertheless, agencies may be willing to serve as
representative payee if a tenant wants that arrangement. Far fewer agencies in 2007 reported
that they offer this service (30 percent) than did so in 2004 (74 percent).
•
Every Los Angeles PSH project identified the same three circumstances that would cause
them to evict a tenant - disruptive or aggressive behavior that threatened other tenants,
destroying property, and nonpayment of rent. There was no change from 2004 to 2007.
Even with these circumstances, projects usually try to work with a tenant to resolve the
behavior, and only evict when it is clear that the tenant will not stop.
•
Very few Los Angeles PSH projects insist on abstinence - only 6 percent in either year said
they would evict for substance use, and not many more (12 percent in 2004 and 6 percent in
2007) said they would evict even after multiple relapses.
•
A substantial, but shrinking, proportion of programs say that they condition tenancy on
service participation - 40 percent of PSH projects in 2007 (down from 56 percent in 2004)
said they would evict for failure to participate in services.
Conclusions
The information supplied by PSH agencies and projects has shed light on some important issues
that may be useful for shaping policy. Before 2004 we did not have any of this information; now we
have it for two different years that straddle the beginning of major efforts to stimulate PSH and
change public policies about how to address chronic homelessness. In particular, we now have
some meaningful idea of agencies’ staying power and intent in the PSH arena, their mission, the
populations they are willing and able to serve, and their policies with respect to preservation of
tenancy. It appears that the large majority of PSH projects and agencies operate within the scope of
what CSH considers appropriate for PSH - serving highly disabled people, not conditioning tenancy
on service participation, and offering a range of supportive services. However, there still remain a
sizable proportion of projects that do tie tenancy to service participation. Most agencies offering
PSH in Los Angeles are “in it for the long haul,” including most of the agencies that were new to
our survey in 2007, and often new to PSH as well.
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
20
EVIDENCE FOR THE EFFECTIVENESS OF HOUSING AND EMPLOYMENT
SERVICES FOR CHRONICALLY HOMELESS PEOPLE WITH MULTIPLE
DISABILITIES
Los Angeles has been home to two demonstration projects supported by the federal government’s
Chronic Homelessness Initiative. The first of these, the Skid Row Collaborative, had the goal of
recruiting and providing stable housing for the hardest-to-serve Skid Row residents. The second,
LA's HOPE, had the same housing goal as the Skid Row Collaborative, but also had the mission to
move tenants into employment. To qualify for either, a person had to be chronically homeless and
have a serious mental illness; most also had a co-occurring substance abuse disorder and many had a
wide range of other health problems. Evaluations are available for both (Burt 2007 for the Skid
Row Collaborative, and Burt 2008b for LA's HOPE. Both show that substantial success may be
achieved with even the hardest-to-serve long-term homeless people with a well-designed project that
has the right kind of resources and enough of them.
The Skid Row Collaborative had three years of federal demonstration funding, which ended in June
2007. LA's HOPE is in its fifth and final year of federal funding, which ends in fall 2008. After
describing the outcomes these demonstrations achieved, we conclude this section of the report with
a discussion of their future, focusing on whether the local public agencies that were partners in the
demonstrations have stepped forward with the resources needed for these projects to continue into
the future.
The Skid Row Collaborative
An evaluation of the Skid Row Collaborative (SRC) completed in spring 2007 (Burt 2007b) focused
primarily on process issues, but was able to use housing retention data for residents of the
Collaborative and of other Skid Row Housing Trust (SRHT) buildings to assess the impact of
participating in the Collaborative on housing retention.
The SRC structure of services integrated with housing has helped to keep small tenant lapses from
becoming major ones. SRC tenants live in the St. George and three or four other nearby SRHT
permanent supportive housing projects. Stationed at the St. George as part of the SRC were a
psychiatric nurse, a psychiatrist and a VA social worker who remained in their positions throughout
the entire demonstration. During the demonstration period the nurse was on-site full time, the
psychiatrist half-time, the VA social worker was available full-time, and these three key staff were
supported by several case managers with offices in the St. George. Some specialty staff were also
sometimes available, such as a post-traumatic stress counseling specialist who joined the staff late in
the demonstration as the need to address PTSD became very apparent. Staff saw tenants regularly.
The nurse dispensed all medications and was thus the first to know that someone had stopped
taking needed meds. She would ask the person to “drop by and see me.” Most of the time this
approach resulted in resumption of medications as well as discussion of what might have been going
on that induced the tenant to think of stopping. If the nurse’s intervention was not enough to
resolve the issue, the nurse and psychiatrist were able to work together to develop an approach that
was almost always successful.
The intensive supervision that results from this program structure has let the SRC staff help even
people with potentially extreme behavior to keep their housing and deal with their issues. One
example was a fire setter who always started out small (such as a piece of paper in her room). The
minute matches started disappearing, the nurse and the psychiatrist knew the tenant was starting to
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
21
go off medications and took steps to get the tenant back on track. It was the staff’s opinion,
including that of the psychiatrist, that the SRC was able to house much sicker people successfully
with the services in place than it could have done with less intense and less integrated services.
Documenting the SRC’s Effectiveness
The primary goal of the SRC was to end the homelessness of chronically homeless, very disabled
people with serious mental illness by providing housing and the supportive services that help them
stay in housing. To assess the SRC’s effectiveness, the evaluator looked at the project’s own
statistics on housing placement and retention and compared these to housing retention among
disabled formerly homeless people living in the Simone and the Pershing (two other SRHT hotels)
with subsidies from Shelter Plus Care housing certificates.
In interpreting the results, it is important for the reader to keep two things in mind - the SRC
tenants were significantly more disabled than tenants in the comparison group, and they were
homeless for considerably longer than the comparison group tenants before they obtained housing.
Simone and Pershing Shelter Plus Care tenants had to be homeless and disabled to qualify for their
housing unit. Their disabilities were HIV/AIDS, substance abuse, or mental illness. In contrast the
SRC tenants had to be chronically homeless (either living on the streets or in an emergency shelter
for over a year, or four episodes of homelessness in three years) and severely mentally ill, with or
without a co-occurring substance abuse disorder. In reality the average length of homelessness for
the SRC tenants was eight years before coming into housing, and almost all had co-occurring
substance abuse disorders. The SRC staff believed the project’s tenants to be much more ill and in
need of much more direct and ongoing support than the Shelter Plus Care tenants in the Simone
and Pershing hotels. As one senior SRHT staff person put it:
We screen people applying for a unit in our hotels. The screening process allows us
to “flag” a person’s circumstances that we know from past experience mean the
person will have trouble retaining housing. Too many flags and we don’t take that
person. For the SRC, everyone recruited had so many flags that we would never
have offered them housing if we had followed our normal procedures. But we took
them all into the SRC….and to my amazement, it has worked. We’ve done better
with harder people. I’ve personally done a 180 on thinking that housing and services
should be co-located.
Thus if the SRC retention patterns are similar to those in the Simone and the Pershing, it means the
SRC is achieving the same results for considerably more difficult clients. If the retention patterns
are better for SRC tenants than for Simone and Pershing tenants, it means the SRC is doing better
with harder clients. Either finding would speak well for the SRC.
Meeting Housing Retention Milestones
The SRC was able to house 62 people at one time, and has housed 101 people between January 1,
2004, the date its first tenant moved in, and May 31, 2007, the cutoff date used in the evaluation.
During that same period 139 new Shelter Plus Care tenants moved into the Simone and 65 new
Shelter Plus Care tenants moved into the Pershing. We treat these 204 tenants as a comparison
group. We established a length of stay for each tenant by calculating the number of days between
his or her move-in date and the date he or she moved out or May 31, 2007, whichever came sooner.
In a few instances a tenant had more than one period of residency with a gap between during which
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
22
he or she was out of housing. When this happened the days of each residency were combined to
calculate a total number of days in housing.
The first way of assessing the SRC’s success in housing retention was to look at the average length
of stay for SRC and comparison group clients. Results indicate that the average length of stay for
SRC clients was 614 days, which was about 200 or 210 days longer than the average length of stay
for Simone and Pershing Shelter Plus Care tenants - close to 7 months longer.
However, this approach fails to take into consideration when each tenant moved in, and therefore
whether he or she had a chance to stay for a long time. A tenant moving in on January 1, 2004 had
a chance to stay in housing for three years and five months, whereas a tenant moving in on January
1, 2007 only had a chance to stay for five months. The most appropriate, and revealing, way to
examine housing retention is to look at those who achieved a specific milestone such as still being in
housing after one year, assuming they were eligible for that milestone. To be eligible for
consideration as having reached the one-year milestone, a tenant would have had to move in at least
one year before May 31, 2007. To be eligible for consideration as having reached the two-year
milestone, a tenant would have had to move in at least two years before May 31, 2007, and so on.
Since the SRC only started a little more than three years before the evaluation took place, the
number of eligible tenants for each milestone got smaller as the milestone involved longer housing
retention.
Figure 4: Housing Retention, SRC
and Comparison Hotels
SRC
Percent of tenants
reaching milestone
Figure 4 displays the results in
graphic form. Table 7 provides
the relevant data, examining the
milestone achievement of
people who moved into SRC
housing or the Shelter Plus
Care-supported housing at the
Simone and Pershing hotels
from January 1, 2004, through
May 31, 2007.
80
70
60
50
40
30
20
10
0
Comparison Hotels
75
63
59
59
38
Table 7 reports the number of
tenants eligible for each
14
milestone and the number and
proportion of those eligible who
reached the milestone. For the
1+ years*
2+ years*
3+ years*
milestone of one-year housing
•Data points represent proportion of tenants reaching milestone, of all those moving into housing early enough
retention, 92 SRC tenants were
that they could have reached milestone (I.e., at least 1 (2, 3) years before May 31, 2007.
eligible, of whom 75 percent (69
tenants) met the milestone. This is significantly higher (p < .05) than the milestone achievement of
people in the comparison hotels, among whom 155 were eligible and 63 percent (97 tenants) met the
milestone. The difference in housing retention increases in favor of SRC tenants at each subsequent
milestone. At the two-year milestone it is 59 percent for SRC tenants versus 38 percent for
comparison hotel tenants (p < .01); at the three-year milestone the difference is 59 versus 14 percent
(p < .001).
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
23
Table 7: Housing Retention for SRC Tenants
Compared to Shelter Plus Care Tenants at the Simone and Pershing Hotels
(Move-in on January 1, 2004 or Thereafter, through May 31, 2007)
Client Group
Skid Row Collaborative Tenants
One or More Years
Two or More Years
Three or More Years
S+C Tenants in Comparison Hotels
One or More Years
Two or More Years
Three or More Years
Number of Eligibles--Tenants
whose Move-in Date was at
least 1, 2, or 3 Years before
5-31-2007
Number of Eligible
Tenants who Reached
the Milestone of 1, 2, or
3 Years in Housing
Proportion of Eligible
Tenants who Reached the
Milestone
92
69
32
69
41
19
75a
59b
59c
155
106
37
97
40
5
63a
38b
14c
a Percentages
so marked are significantly different from each other at p < .05.
so marked are significantly different from each other at p < .01.
c Percentages so marked are significantly different from each other at p < .001.
b Percentages
The level of staffing committed to helping SRC tenants retain housing was considerable. The SRC,
staff-to-tenant ratio was 1 to 20. Case management ratios for the Simone and the Pershing were
1:82 and 1:67, respectively, and the specialized SRC staff (psychiatrist, nurse, VA specialist) were not
available. One might look at the data in table 7 for one-year retention and think that performance of
the Simone and Pershing was really not too bad. Considering the significantly lower availability of
supportive services, their retention rate for the one-year milestone is only 12 percent lower than that
achieved by the SRC (63 vs. 75 percent). However, the rates of those eligible for and reaching the
two- and three-year milestones tell a different story. For those eligible for the three-year milestone,
SRC retention is stable at 59 percent while retention at the comparison hotels has fallen to 14
percent. If both the SRC and Shelter Plus Care are programs intended to provide permanent
supportive housing and to assure that people have the ability to remain in it, then the SRC model is
clearly preferable. 5
Also examined was the issue of differences within the SRC, specifically between tenants housed in
the St. George Hotel where service staff have offices within the building and those housed
elsewhere but served by the St. George-based support staff. That is, does being in the same building
as supportive services staff do more for tenants than just having those staff available in a nearby
location? The answer is “no” - there are no significant differences in housing retention between
SRC participants housed in the St. George and those housed elsewhere, after taking into account
each tenant’s eligibility for reaching a retention milestone.
Tenants in both groups had episodes in the early months after move-in that threatened their
continued residence; likewise those still in housing after a year or so faced possibilities of relapse,
decompensation, and the like. The difference that made a difference was that SRC staff, with a ratio
5
We do not have information about where people in SRC or in the comparison hotels went after leaving that
housing. In stating that the SRC “did better,” we are assuming that there are no differences in the proportions from
each group that moved to equivalent or better accommodations or to less good situations or back to homelessness.
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
24
of 1 caseworker to 20 residents and the three on-site professional staff were in a position to observe
the beginnings of such episodes and take steps to avert them, whereas staff at the Pershing and
Simone, with caseworker ratios three and four times as high and no on-site psychiatric staff, could
not do the same. The result was that comparison group tenants, although they may have had less
intense or complex problems than SRC tenants, nevertheless lost housing at a higher rate.
LA's HOPE
LA’s HOPE has the very concrete goals of housing its clients and helping them get and keep
employment. The LA's HOPE demonstration is able to house 76 clients at one time using Shelter
Plus Care certificates that came as part of the project’s federal grant (they are renewable during
future funding cycles in the same manner as for other Shelter Plus Care grants, so they represent net
new units of PSH for Los Angeles). LA’s HOPE had served 147 clients between its first intake in
late winter 2003 and May 2007, the final month that could be included in the outcomes data to be
analyzed for this evaluation.
LA's HOPE provides housing subsidies through the city’s housing authority (HACLA) and delivers
services through the workforce development (CDD) and mental health (DMH) systems. The latter
two systems in turn have both policy and practice components - the city’s Community Development
Department and some of its One-Stops (known locally as WorkSource Career Centers), and the
county’s Department of Mental Health and three of the mental health agencies under contract to
DMH that offer a special state funded program known as AB 2034. AB 2034 programs have the
explicit goal of preventing or ending homelessness for people with serious mental illness (for a
detailed description, see Burt and Anderson 2005). The three AB 2034 programs chosen to
participate in LA's HOPE received extra AB 2034 resources to provide supportive services to LA's
HOPE clients - that is, their AB 2034 caseloads increased by the number of LA's HOPE clients
assigned to their program. LA's HOPE clients thus received “usual” AB 2034 services and also had
access to Shelter Plus Care certificates reserved for LA's HOPE participants and extensive supports
to prepare for, obtain, and keep a job.
A potential participant in LA's HOPE is first recruited and enrolled by an AB 2034 program, whose
staff assist the person to find housing and offer other needed supportive and mental health services.
Once housed, the person is expected to begin participating in employment-related activities. While
enrollees did not have to sign an agreement to participate in employment, recruitment procedures
clearly indicated that the purpose of the program was to help people to move toward employment
and potential enrollees were screened for their stated motivation to work with the program to that
end. Work-related activities usually began with work experience jobs in the mental health programs
and then moved on to full- or part-time employment in the competitive job market. One-Stops
were instrumental in helping LA's HOPE clients get some jobs of the latter type, but for the most
part clients found their jobs on their own or with the help of LA's HOPE case managers.
Housing and Employment Outcomes of LA's HOPE
Because LA's HOPE participants are also AB 2034 clients and consistent information about clients
of every AB 2034 program statewide is reported to a central database every month, evaluators were
able to construct comparison groups of other AB 2034 clients in Los Angeles County to assess the
success of LA's HOPE in getting its clients into permanent housing and employment. The results
were impressive (see Burt 2008b for details):
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
25
•
LA's HOPE clients were significantly more likely than other AB 2034 clients to engage in
full-time employment, part-time employment, and competitive employment, and
significantly less likely to have avoided employment altogether.
•
LA's HOPE clients spent more than twice as many days in all types of employment than did
other AB 2034 clients, and more than four times the days working in competitive
employment.
•
LA's HOPE clients spent many more days in PSH than did other AB 2034 clients, and far
fewer days incarcerated or hospitalized.
•
Multivariate analyses showed program differences (LA's HOPE vs. other AB 2034 clients) to
be the only significant factor consistently affecting the employment outcomes. For housing,
program differences were the strongest consistent factor. In addition, holding group
membership constant, some personal characteristics consistently affected housing outcomes.
o Compared to white AB 2034 clients, African-American and Hispanic clients were only
about one-fourth as likely to have lived in PSH since enrollment, took about three times
as long to move into PSH (for those who did make that move), and spent fewer days in
PSH since enrollment.
o AB 2034 clients with a co-occurring substance abuse disorder took only half the time to
move into PSH as those without such a disorder and had spent an average of 10 or 11
more days in PSH since enrollment.
The Future of the Skid Row Collaborative and LA's HOPE
The assumption underlying the federal Chronic Homelessness Initiative demonstrations was that
they would show their value to the local public agencies that participated in the projects as partners,
to the extent that the public agencies would be willing, and able, to cover the costs of their
continuation once federal funding expired. HACLA, the agency responsible for the housing subsidy
component of both projects, is committed to continuing its involvement, but expectations that the
agencies offering supportive services will devote local resources to maintaining these services have
for the most part not been fulfilled.
For both demonstrations, the only secure ongoing financial commitments have come from HACLA,
through which the projects’ Shelter Plus Care certificates have run. Those certificates remain active
after the other federal funding sources stop, so they continue to subsidize the rents of the people
who were in both demonstrations. HACLA is also committed to keeping those certificates available
to new recipients should the current recipients leave their housing and other people appropriate to
the projects be selected to receive them.
DMH, which supplied a half-time psychiatrist for the SRC and AB 2034 funding for LA's HOPE
has run into enormous budget crises of its own. It has not renewed its commitment to provide
psychiatric care for the SRC, but it has been working with Lamp Community, the primary SRC
service provider, to help Lamp qualify as an agency that can bill to Medi-Cal, and has allowed Lamp
to use some of its DMH contract funds as the county match for such billing. These arrangements
do not immediately benefit the SRC tenants (they only help other Lamp clients), but the additional
resources generated by Medi-Cal billing may ultimately be useful in supporting these tenants. To
help Lamp continue to support the SRC tenants until these arrangements begin to bear fruit, CSH
has contributed about half a million dollars in grant funds it received from The California
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
26
Endowment to sustain this collaborative until Medi-Cal funding can begin to have an effect. CSH
has also provided extensive technical assistance to help collaborative members increase their
capacity to bill Medi-Cal for ongoing services.
With respect to LA's HOPE, Governor Schwarzenegger’s line item veto in fall 2007 of state funding
for the AB 2034 program has thrown the future of that program into grave doubt. No new clients
are being accepted by any AB 2034 program anywhere in the state unless local funds can make up
for the disappearing state dollars. In Los Angeles County DMH had a major budget shortfall even
before the demise of AB 2034 funding, and is not in any position to compensate. Thus it is very
unclear at this time what level of mental health care and supports will be available to existing AB
2034 clients (including LA's HOPE clients) after this year, and any expectation that new clients
might be enrolled once the demonstration ends are likely to be disappointed.
Each demonstration has one more public partner that has made some commitment to ongoing
activity. For the SRC, that partner is the VA, which is continuing to support the social worker who
was part of the on-site SRC services team from the beginning. For LA's HOPE, the final player is
CDD, which was responsible for managing the demonstration. CDD expects to continue to
encourage its One-Stops to partner with mental health agencies to be able to improve its services to
customers with mental disabilities. But it will not have the resources to support specialized
employment case managers whose job it is to work solely with LA's HOPE clients or similar
homeless or formerly homeless people. In addition, many of the mental health agencies with which
the One-Stops might partner will be feeling the effects of losing AB 2034 funds and may not be in a
position to pursue innovative arrangements with other agencies for some time to come.
CONCLUSIONS
There is definitely some good news in this report related to major goals for Los Angeles of CSH’s
Hilton Foundation Project to End Homelessness among People with Mental Illness. Permanent
supportive housing development is in full swing, with about 600 new units already open and at least
another 2,200 in the pipeline. New projects appear to be targeting chronically homeless people
specifically, which will have the effect of moving more people with serious mental illness into
housing. Federal demonstration projects have contributed solid evidence that PSH with substantial
service levels and innovative approaches are very effective strategies to ending the homelessness of
people with many disabilities who have not responded to other program structures or who quickly
fail in them because they do not receive the types of attention they need. Philanthropy and local
government are collaborating to help extend the life of these federal demonstration projects once
the federal resources are withdrawn, but it remains to be seen whether the models will be adopted as
local practice, with local resources devoted to them, despite their clear ability to achieve their goals.
The sheer numbers of homeless and chronically homeless people in the county appear to be down,
although some proportion of the decrease may be due to methodological changes in counting and
interviewing procedures. All of these findings are in the direction that the Hilton Foundation
Project is working to move the county.
The first evaluation report for this project (Burt 2005a) covered the issues of numbers (how many
homeless people are there?); baseline estimates of available PSH; characteristics of PSH tenants in
2004; and sources and amounts of capital, operating, and services funding for projects open in 2004.
For these issues that first report relied on the 2005 LAHSA point-in-time count of homeless people
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
27
in the county (LAHSA 2005) and the report done for THCH on the 2004 surveys of agencies and
projects offering PSH (Burt 2005b).
That first report also contained a brief assessment of the policy environment and baseline
organizational activities relating to ending long-term homelessness and PSH development. The
present report does not update this information because separate reports are now being written that
cover the policy environment and progress toward developing policies and resources that will help
to end homelessness. The second report for this evaluation (Burt 2007a) presented this analysis of
the policy developments and changes that had occurred between early 2005 (when the Hilton
Foundation Project began) and early 2007. The fourth report for this evaluation, scheduled for
summer 2008, will repeat that policy analysis, as will the final report, due in early 2010 and reflecting
circumstances in 2009 after five years of project activity.
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
28
REFERENCES
Burt, M.R. 2008a. Evolution of PSH in Taking Health Care Home Communities, 2004–2007:
Tenants, Programs, Policies, and Funding at Project End. Oakland, CA: Corporation for
Supportive Housing. Available at www.csh.org/resources/publications.
Burt, M.R. 2008b. Evaluation of LA’s HOPE: Ending Chronic Homelessness Through
Employment and Housing, Final Report. Washington, DC: The Urban Institute. Available
at www.urban.org.
Burt, M.R. 2008c. Pushing the Envelope: Deepening and Broadening Involvement in THCH
Communities as Projects End. Oakland, CA: Corporation for Supportive Housing.
Available at www.csh.org/resources/publications.
Burt, Martha R. 2007a. Hilton Foundation Project to Reduce Homelessness Among People with
Serious Mental Illness: System Change Efforts and Their Results, 2005-2006. Oakland, CA:
Corporation for Supportive Housing (second evaluation report). Available at
http://www.urban.org/url.cfm?ID=411449.
Burt, M.R. 2007b. The Skid Row Collaborative—2003–2007: Process Evaluation. Washington,
DC: The Urban Institute. Available at http://www.urban.org/url.cfm?ID=411546.
Burt, M.R. 2005a. Hilton Foundation Project To End Homelessness among People With Serious
Mental Illness: First Evaluation Report. Oakland, CA: Corporation for Supportive Housing.
Burt, M.R. 2005b. Taking Health Care Home: Baseline Report on Tenants, Programs, Policies, and
Funding. Oakland, CA: Corporation for Supportive Housing. Available at
www.csh.org/resources/publications.
Burt, M.R. and Anderson, J. 2006. Taking Health Care Home: Impact of System Change Efforts at
the Two-Year Mark. Oakland, CA: Corporation for Supportive Housing. Available at
www.csh.org/resources/publications.
Burt, M.R. and Anderson, J. 2005. AB2034 Program Experiences in Housing Homeless People
with Serious Mental Illness. Oakland, CA: Corporation for Supportive Housing.
Los Angeles Homeless Services Authority. 2007. Greater Los Angeles Homeless Count. Accessed
at www.lahsa.org on October 15, 2007
Los Angeles Homeless Services Authority. 2005. 2005 Greater Los Angeles Homeless Count.
Accessed at www.lahsa.org on January 8, 2007.
Corporation for Supportive Housing:
Hilton Foundation Project to End Homelessness for People with Mental Illness in Los Angeles
29